import pandas as pd
import plotly.express as px
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt
from matplotlib.dates import MonthLocator, num2date
from matplotlib.ticker import FuncFormatter
import datetime
import os
import warnings
warnings.filterwarnings('ignore')
%matplotlib inline
pd.options.display.float_format = '{:.2f}'.format
from scipy.stats import skew,kurtosis,zscore
import matplotlib

sns.set_style('darkgrid')
matplotlib.rcParams['font.size'] = 14
matplotlib.rcParams['figure.figsize'] = (10, 6)
matplotlib.rcParams['figure.facecolor'] = '#00000000'

os.chdir('C:\\Users\\lance\\OneDrive\\Desktop\\DATA science\\Work_Hours\\')

df_train=pd.read_csv('train.csv')
df_test=pd.read_csv('test.csv')

df_train.head(10)

	age	workclass	fnlwgt	education	education-num	marital-status	occupation	relationship	race	sex	capital-gain	capital-loss	hours-per-week	native-country	income
0	52	Private	98926	HS-grad	9	Never-married	Other-service	Unmarried	White	Male	-3	0	44	United-States	<=50K
1	76	Self-emp-not-inc	132607	Bachelors	12	Married-civ-spouse	Sales	Husband	White	Male	7527	0	44	United-States	>50K
2	40	Private	243258	Bachelors	13	Married-civ-spouse	Transport-moving	Husband	Amer-Indian-Eskimo	Male	19	0	57	United-States	>50K
3	76	State-gov	181259	Bachelors	12	Married-civ-spouse	Prof-specialty	Husband	White	Male	7720	0	44	United-States	>50K
4	36	Self-emp-inc	115379	Masters	13	Divorced	Exec-managerial	Not-in-family	White	Male	12	0	58	United-States	<=50K
5	24	Private	274505	Bachelors	12	Never-married	Prof-specialty	Own-child	White	Female	2	0	40	United-States	<=50K
6	40	Private	80789	HS-grad	8	Married-civ-spouse	Adm-clerical	Wife	White	Female	13	0	35	United-States	>50K
7	42	Federal-gov	243759	Some-college	10	Divorced	Adm-clerical	Other-relative	Asian-Pac-Islander	Female	18	0	39	Cuba	>50K
8	37	Private	67638	Bachelors	12	Married-civ-spouse	Exec-managerial	Husband	Asian-Pac-Islander	Male	4	0	39	Philippines	>50K
9	52	Self-emp-not-inc	130237	HS-grad	8	Married-civ-spouse	Sales	Husband	White	Male	0	1	55	Mexico	<=50K

df_train.describe()

	age	fnlwgt	education-num	capital-gain	capital-loss	hours-per-week
count	18944.00	18944.00	18944.00	18944.00	18944.00	18944.00
mean	39.27	164292.04	9.48	577.48	33.08	40.12
std	13.40	93425.99	2.83	2414.47	248.52	11.12
min	14.00	8019.00	2.00	-58.00	-2.00	1.00
25%	28.00	98530.75	8.00	-5.00	0.00	35.00
50%	40.00	152895.50	9.00	5.00	0.00	39.00
75%	47.00	215541.00	12.00	17.00	0.00	44.00
max	99.00	772988.00	16.00	16193.00	2437.00	102.00

df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18944 entries, 0 to 18943
Data columns (total 15 columns):
 #   Column          Non-Null Count  Dtype 
---  ------          --------------  ----- 
 0   age             18944 non-null  int64 
 1   workclass       18944 non-null  object
 2   fnlwgt          18944 non-null  int64 
 3   education       18944 non-null  object
 4   education-num   18944 non-null  int64 
 5   marital-status  18944 non-null  object
 6   occupation      18944 non-null  object
 7   relationship    18944 non-null  object
 8   race            18944 non-null  object
 9   sex             18944 non-null  object
 10  capital-gain    18944 non-null  int64 
 11  capital-loss    18944 non-null  int64 
 12  hours-per-week  18944 non-null  int64 
 13  native-country  18944 non-null  object
 14  income          18944 non-null  object
dtypes: int64(6), object(9)
memory usage: 2.2+ MB

df_train.shape

(18944, 15)

df_train.workclass.value_counts()

 Private             11700
 Self-emp-not-inc     3089
 Local-gov            1159
 Federal-gov           995
 Self-emp-inc          861
 State-gov             819
 Without-pay           258
 Never-worked           63
Name: workclass, dtype: int64

px.histogram(df_train,x='workclass',y='hours-per-week')

for column in df_train.columns:
    print(f"{column} = {df_train[df_train[column] == '?'].shape[0]}")

age = 0
workclass = 0
fnlwgt = 0
education = 0
education-num = 0
marital-status = 0
occupation = 0
relationship = 0
race = 0
sex = 0
capital-gain = 0
capital-loss = 0
hours-per-week = 0
native-country = 0
income = 0

df_train.education.value_counts()

 HS-grad         5462
 Some-college    3745
 Bachelors       3337
 11th            1000
 Masters          941
 Assoc-acdm       796
 Assoc-voc        769
 7th-8th          750
 Prof-school      476
 Doctorate        387
 10th             386
 5th-6th          319
 9th              200
 12th             184
 1st-4th          148
 Preschool         44
Name: education, dtype: int64

fig = px.histogram(df_train, 
                   x='age', 
                   marginal='box', 
                   nbins=62, 
                   title='Distribution of Hrs per week')
fig.update_layout(bargap=0.1)
fig.show()

The above histogram shows many people in their 40's and 50's are enrolled in the work force.A further analysis as to their occupation will throw some light on the same

fig = px.histogram(df_train, 
                   x='hours-per-week', 
                   marginal='box', 
                   nbins=86, 
                   title='Distribution of Age')
fig.update_layout(bargap=0.1)
fig.show()

The above histogram shows that most people work approx 40 hrs a week

fig = px.histogram(df_train, 
                   x='hours-per-week', 
                   marginal='box', 
                   nbins=86, 
                   color='sex', 
                   color_discrete_sequence=['green', 'grey'],
                   title='Distribution Hrs per week')
fig.update_layout(bargap=0.1)
fig.show()
                  

df_train.income.value_counts()

 <=50K    13517
 >50K      5427
Name: income, dtype: int64

Analysis of age vs income

px.scatter(df_train,x='age',y='hours-per-week',color='sex',title='Age vs Hrs per week')

Education vs Hrs per week

px.histogram(df_train,x='education',y='hours-per-week',color='sex',title='Education vs Hrs per week')

Occupation vs Hrs per week

px.histogram(df_train,x='occupation',y='hours-per-week',color='sex',title='Occupation vs Hrs per week')

Country vs Hrs week

df_train['native-country'].value_counts()

 United-States                 16000
 South                           342
 Mexico                          329
 Cuba                            321
 Puerto-Rico                     283
 Jamaica                         254
 Honduras                        247
 Canada                          236
 England                         165
 India                           125
 Philippines                     112
 Taiwan                           49
 Dominican-Republic               48
 Vietnam                          42
 Germany                          34
 China                            29
 Haiti                            27
 Poland                           26
 Italy                            24
 Guatemala                        24
 Japan                            22
 El-Salvador                      22
 Columbia                         19
 Laos                             19
 Iran                             17
 Thailand                         15
 Yugoslavia                       13
 Peru                             12
 Greece                           12
 Portugal                         12
 Nicaragua                        11
 Trinadad&Tobago                   9
 Ireland                           9
 Ecuador                           7
 Cambodia                          7
 Outlying-US(Guam-USVI-etc)        6
 Scotland                          5
 Hungary                           4
 Hong                              3
 Holand-Netherlands                2
 France                            1
Name: native-country, dtype: int64

df_train['race'].value_counts()

 White                 12682
 Black                  3565
 Asian-Pac-Islander     1728
 Amer-Indian-Eskimo      491
 Other                   478
Name: race, dtype: int64

Race Vs Hrs per week

px.histogram(df_train,x='race',y='hours-per-week',color='sex',title='Race vs Hrs per week')

cor=df_train.corr()
cor

	age	fnlwgt	education-num	capital-gain	capital-loss	hours-per-week
age	1.00	-0.11	0.07	0.09	0.02	0.20
fnlwgt	-0.11	1.00	-0.11	-0.02	-0.01	-0.08
education-num	0.07	-0.11	1.00	0.14	0.05	0.21
capital-gain	0.09	-0.02	0.14	1.00	0.04	0.11
capital-loss	0.02	-0.01	0.05	0.04	1.00	0.04
hours-per-week	0.20	-0.08	0.21	0.11	0.04	1.00

sns.heatmap(cor,annot=True);

numeric_cols=(df_train.select_dtypes(include=['int64']))
numeric_cols

	age	fnlwgt	education-num	capital-gain	capital-loss	hours-per-week
0	52	98926	9	-3	0	44
1	76	132607	12	7527	0	44
2	40	243258	13	19	0	57
3	76	181259	12	7720	0	44
4	36	115379	13	12	0	58
...	...	...	...	...	...	...
18939	25	174301	12	5	0	39
18940	22	135915	9	16	0	44
18941	47	47276	12	1	1842	39
18942	48	88787	12	-9	0	71
18943	47	175720	13	-1	0	39

18944 rows × 6 columns

list(numeric_cols)

['age',
 'fnlwgt',
 'education-num',
 'capital-gain',
 'capital-loss',
 'hours-per-week']

sns.boxplot(data=numeric_cols);

sns.catplot(y="education", x="hours-per-week", data=df_train);

sns.catplot(y="occupation", x="hours-per-week", data=df_train);

sns.boxplot(y="education", x="hours-per-week",hue='sex' ,data=df_train);

sns.boxplot(y="occupation", x="hours-per-week", data=df_train);

sns.boxplot(y="marital-status", x="hours-per-week",hue='sex', data=df_train);

df_train.hist();

fig,axes=plt.subplots(2,3,figsize=(25,15))

axes[0,0].set_title('Age gain vs hours per week')
sns.scatterplot(y="age", x="hours-per-week",hue='sex',ax=axes[0,0], data=df_train);



axes[0,1].set_title('Education vs hours per week')
sns.violinplot(y="education", x="hours-per-week",hue='sex',ax=axes[0,1], data=df_train);

axes[0,2].set_title('Occupation vs hours per week')
sns.violinplot(y="occupation", x="hours-per-week",hue='sex',ax=axes[0,2], data=df_train);

axes[1,0].set_title('Marital vs hours per week')
sns.violinplot(y="marital-status", x="hours-per-week",hue='sex',ax=axes[1,0], data=df_train);

axes[1,1].set_title('Capital loss vs hours per week')
sns.scatterplot(y="capital-loss", x="hours-per-week",hue='sex',ax=axes[1,1], data=df_train);


axes[1,2].set_title('Capital -gain vs hours per week')
sns.scatterplot(y="capital-gain", x="hours-per-week",hue='sex',ax=axes[1,2] ,data=df_train);

plt.tight_layout(pad=3);

def distribution_check(df,Cols):
    plt.figure(figsize=(25,7))
    fig = 1
    i = (len(Cols)//3)+1
    for col in Cols:
        sk = " Skewness : " + str(skew(df_train[col])) +"\nKurtosis :" + str(kurtosis(df_train[col]))
        plt.subplot(i, 3, fig)
        sns.distplot(df[col]).set_title(sk)
        fig = fig+1

distribution_check(df_train,['age','hours-per-week','capital-gain'])

pip install scipy

Requirement already satisfied: scipy in c:\users\lance\anaconda3\lib\site-packages (1.7.1)
Requirement already satisfied: numpy<1.23.0,>=1.16.5 in c:\users\lance\anaconda3\lib\site-packages (from scipy) (1.20.3)
Note: you may need to restart the kernel to use updated packages.

from scipy import stats 

numeric_cols.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18944 entries, 0 to 18943
Data columns (total 6 columns):
 #   Column          Non-Null Count  Dtype
---  ------          --------------  -----
 0   age             18944 non-null  int64
 1   fnlwgt          18944 non-null  int64
 2   education-num   18944 non-null  int64
 3   capital-gain    18944 non-null  int64
 4   capital-loss    18944 non-null  int64
 5   hours-per-week  18944 non-null  int64
dtypes: int64(6)
memory usage: 888.1 KB

#removing Outliars
z_scores = stats. zscore(numeric_cols) 
abs_z_scores = np. abs(z_scores)
filtered_entries = (abs_z_scores < 3). all(axis=1)
new_df = numeric_cols[filtered_entries]
print(new_df)

       age  fnlwgt  education-num  capital-gain  capital-loss  hours-per-week
0       52   98926              9            -3             0              44
1       76  132607             12          7527             0              44
2       40  243258             13            19             0              57
3       76  181259             12          7720             0              44
4       36  115379             13            12             0              58
...    ...     ...            ...           ...           ...             ...
18938   24  106178             10             2             0              35
18939   25  174301             12             5             0              39
18940   22  135915              9            16             0              44
18942   48   88787             12            -9             0              71
18943   47  175720             13            -1             0              39

[17722 rows x 6 columns]

df_train.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 18944 entries, 0 to 18943
Data columns (total 15 columns):
 #   Column          Non-Null Count  Dtype 
---  ------          --------------  ----- 
 0   age             18944 non-null  int64 
 1   workclass       18944 non-null  object
 2   fnlwgt          18944 non-null  int64 
 3   education       18944 non-null  object
 4   education-num   18944 non-null  int64 
 5   marital-status  18944 non-null  object
 6   occupation      18944 non-null  object
 7   relationship    18944 non-null  object
 8   race            18944 non-null  object
 9   sex             18944 non-null  object
 10  capital-gain    18944 non-null  int64 
 11  capital-loss    18944 non-null  int64 
 12  hours-per-week  18944 non-null  int64 
 13  native-country  18944 non-null  object
 14  income          18944 non-null  object
dtypes: int64(6), object(9)
memory usage: 2.2+ MB

#Check for Outliars using boxplot

sns.boxplot(data=new_df);

dataset=df_train.copy()   
dataset

	age	workclass	fnlwgt	education	education-num	marital-status	occupation	relationship	race	sex	capital-gain	capital-loss	hours-per-week	native-country	income
0	52	Private	98926	HS-grad	9	Never-married	Other-service	Unmarried	White	Male	-3	0	44	United-States	<=50K
1	76	Self-emp-not-inc	132607	Bachelors	12	Married-civ-spouse	Sales	Husband	White	Male	7527	0	44	United-States	>50K
2	40	Private	243258	Bachelors	13	Married-civ-spouse	Transport-moving	Husband	Amer-Indian-Eskimo	Male	19	0	57	United-States	>50K
3	76	State-gov	181259	Bachelors	12	Married-civ-spouse	Prof-specialty	Husband	White	Male	7720	0	44	United-States	>50K
4	36	Self-emp-inc	115379	Masters	13	Divorced	Exec-managerial	Not-in-family	White	Male	12	0	58	United-States	<=50K
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
18939	25	Private	174301	Bachelors	12	Never-married	Other-service	Not-in-family	White	Female	5	0	39	United-States	<=50K
18940	22	Private	135915	Some-college	9	Never-married	Other-service	Unmarried	White	Female	16	0	44	Canada	<=50K
18941	47	Local-gov	47276	Bachelors	12	Never-married	Adm-clerical	Other-relative	Black	Female	1	1842	39	United-States	<=50K
18942	48	Private	88787	Bachelors	12	Married-civ-spouse	Sales	Husband	Asian-Pac-Islander	Male	-9	0	71	South	<=50K
18943	47	Federal-gov	175720	Masters	13	Married-civ-spouse	Transport-moving	Husband	White	Male	-1	0	39	United-States	>50K

18944 rows × 15 columns

from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
dataset[['sex','income']] = dataset[['sex','income']].apply(LabelEncoder().fit_transform)
dataset[['sex','income']] 

	sex	income
0	1	0
1	1	1
2	1	1
3	1	1
4	1	0
...	...	...
18939	0	0
18940	0	0
18941	0	0
18942	1	0
18943	1	1

18944 rows × 2 columns

dataset['marital-status'].unique()

array([' Never-married', ' Married-civ-spouse', ' Divorced',
       ' Married-spouse-absent', ' Separated', ' Widowed',
       ' Married-AF-spouse'], dtype=object)

#1) used the map function
dataset["marital-status"] = dataset["marital-status"].replace([' Never-married',' Divorced',' Separated',' Widowed'],0)
dataset["marital-status"] = dataset["marital-status"].replace([' Married-AF-spouse',' Married-civ-spouse', ' Married-spouse-absent'],1)
display(dataset)

	age	workclass	fnlwgt	education	education-num	marital-status	occupation	relationship	race	sex	capital-gain	capital-loss	hours-per-week	native-country	income
0	52	Private	98926	HS-grad	9	0	Other-service	Unmarried	White	1	-3	0	44	United-States	0
1	76	Self-emp-not-inc	132607	Bachelors	12	1	Sales	Husband	White	1	7527	0	44	United-States	1
2	40	Private	243258	Bachelors	13	1	Transport-moving	Husband	Amer-Indian-Eskimo	1	19	0	57	United-States	1
3	76	State-gov	181259	Bachelors	12	1	Prof-specialty	Husband	White	1	7720	0	44	United-States	1
4	36	Self-emp-inc	115379	Masters	13	0	Exec-managerial	Not-in-family	White	1	12	0	58	United-States	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
18939	25	Private	174301	Bachelors	12	0	Other-service	Not-in-family	White	0	5	0	39	United-States	0
18940	22	Private	135915	Some-college	9	0	Other-service	Unmarried	White	0	16	0	44	Canada	0
18941	47	Local-gov	47276	Bachelors	12	0	Adm-clerical	Other-relative	Black	0	1	1842	39	United-States	0
18942	48	Private	88787	Bachelors	12	1	Sales	Husband	Asian-Pac-Islander	1	-9	0	71	South	0
18943	47	Federal-gov	175720	Masters	13	1	Transport-moving	Husband	White	1	-1	0	39	United-States	1

18944 rows × 15 columns

dataset["marital-status"]=dataset["marital-status"].astype(int)
dataset.dtypes

age                int64
workclass         object
fnlwgt             int64
education         object
education-num      int64
marital-status     int32
occupation        object
relationship      object
race              object
sex                int32
capital-gain       int64
capital-loss       int64
hours-per-week     int64
native-country    object
income             int32
dtype: object

dataset.drop(labels= ['workclass','education','occupation','relationship','race','native-country'],axis =1, inplace = True)

y = dataset['hours-per-week']

y

0        44
1        44
2        57
3        44
4        58
         ..
18939    39
18940    44
18941    39
18942    71
18943    39
Name: hours-per-week, Length: 18944, dtype: int64

x= dataset.drop('hours-per-week',axis=1)
display(x )

	age	fnlwgt	education-num	marital-status	sex	capital-gain	capital-loss	income
0	52	98926	9	0	1	-3	0	0
1	76	132607	12	1	1	7527	0	1
2	40	243258	13	1	1	19	0	1
3	76	181259	12	1	1	7720	0	1
4	36	115379	13	0	1	12	0	0
...	...	...	...	...	...	...	...	...
18939	25	174301	12	0	0	5	0	0
18940	22	135915	9	0	0	16	0	0
18941	47	47276	12	0	0	1	1842	0
18942	48	88787	12	1	1	-9	0	0
18943	47	175720	13	1	1	-1	0	1

18944 rows × 8 columns

from sklearn.model_selection import train_test_split
x_train,x_test,y_train,y_test = train_test_split (x,y , test_size = 0.2, random_state= 42)
print (x_train.shape,x_test.shape,y_train.shape,y_test.shape)

(15155, 8) (3789, 8) (15155,) (3789,)

from sklearn.preprocessing import MinMaxScaler
scaler_train = MinMaxScaler()
scaler_test = MinMaxScaler()

scaler_train.fit(x_train)   # fit only on train dataset
x_train = scaler_train.fit_transform(x_train)

scaler_test.fit(x_test)   # fit only on test dataset
x_test= scaler_test.transform(x_test)

import statsmodels.api as sm

model = sm.OLS(y_train, x_train)
results = model.fit()
print(results.summary())

                                 OLS Regression Results                                
=======================================================================================
Dep. Variable:         hours-per-week   R-squared (uncentered):                   0.911
Model:                            OLS   Adj. R-squared (uncentered):              0.911
Method:                 Least Squares   F-statistic:                          1.939e+04
Date:                Fri, 01 Oct 2021   Prob (F-statistic):                        0.00
Time:                        16:54:55   Log-Likelihood:                         -59701.
No. Observations:               15155   AIC:                                  1.194e+05
Df Residuals:                   15147   BIC:                                  1.195e+05
Df Model:                           8                                                  
Covariance Type:            nonrobust                                                  
==============================================================================
                 coef    std err          t      P>|t|      [0.025      0.975]
------------------------------------------------------------------------------
x1            30.2464      0.620     48.762      0.000      29.031      31.462
x2            26.9356      0.710     37.925      0.000      25.543      28.328
x3            34.7206      0.383     90.607      0.000      33.969      35.472
x4             1.4615      0.250      5.840      0.000       0.971       1.952
x5             8.7765      0.220     39.884      0.000       8.345       9.208
x6             0.3830      0.703      0.545      0.586      -0.994       1.760
x7             0.4712      0.985      0.478      0.632      -1.460       2.402
x8            -1.1269      0.262     -4.300      0.000      -1.641      -0.613
==============================================================================
Omnibus:                      352.617   Durbin-Watson:                   1.962
Prob(Omnibus):                  0.000   Jarque-Bera (JB):              804.278
Skew:                           0.055   Prob(JB):                    2.26e-175
Kurtosis:                       4.123   Cond. No.                         11.4
==============================================================================

Notes:
[1] R² is computed without centering (uncentered) since the model does not contain a constant.
[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.

results.predict(x_test)

array([46.01492359, 47.49218522, 48.35921031, ..., 27.95762458,
       36.72416048, 36.66448297])

y_test

15024    39
11771    43
16615    42
4201     39
13480    43
         ..
8404     39
5412     56
3755     36
11886    34
13592    17
Name: hours-per-week, Length: 3789, dtype: int64

df_test

	age	workclass	fnlwgt	education	education-num	marital-status	occupation	relationship	race	sex	capital-gain	capital-loss	native-country	income	hours-per-week
0	40	Without-pay	106899	HS-grad	8	Never-married	Adm-clerical	Own-child	White	Male	0	0	United-States	<=50K	NaN
1	38	Private	157191	HS-grad	8	Married-civ-spouse	Craft-repair	Husband	Black	Male	0	0	United-States	<=50K	NaN
2	76	Self-emp-not-inc	102547	Some-college	10	Widowed	Sales	Not-in-family	Asian-Pac-Islander	Female	32	0	United-States	<=50K	NaN
3	38	Self-emp-not-inc	44426	HS-grad	9	Never-married	Sales	Unmarried	White	Male	30	0	United-States	>50K	NaN
4	62	Without-pay	33230	HS-grad	9	Married-civ-spouse	Sales	Husband	White	Male	-4	0	United-States	<=50K	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
8114	30	Private	298018	Bachelors	12	Never-married	Adm-clerical	Other-relative	White	Female	20	0	Mexico	<=50K	NaN
8115	40	Private	186841	10th	6	Never-married	Handlers-cleaners	Unmarried	White	Male	0	0	United-States	<=50K	NaN
8116	44	Private	116414	1st-4th	3	Married-civ-spouse	Craft-repair	Wife	Other	Female	10	0	Honduras	<=50K	NaN
8117	41	Federal-gov	185089	Bachelors	12	Never-married	Machine-op-inspct	Not-in-family	Black	Female	1	0	United-States	<=50K	NaN
8118	28	Private	27152	Bachelors	12	Never-married	Exec-managerial	Not-in-family	White	Female	21	0	United-States	<=50K	NaN

8119 rows × 15 columns

df_test.drop('hours-per-week',axis=1,inplace=True)

df_test["marital-status"] = df_test["marital-status"].replace([' Never-married',' Divorced',' Separated',' Widowed'],0)
df_test["marital-status"] = df_test["marital-status"].replace([' Married-AF-spouse',' Married-civ-spouse', ' Married-spouse-absent'],1)
display(df_test)

	age	workclass	fnlwgt	education	education-num	marital-status	occupation	relationship	race	sex	capital-gain	capital-loss	native-country	income
0	40	Without-pay	106899	HS-grad	8	0	Adm-clerical	Own-child	White	Male	0	0	United-States	<=50K
1	38	Private	157191	HS-grad	8	1	Craft-repair	Husband	Black	Male	0	0	United-States	<=50K
2	76	Self-emp-not-inc	102547	Some-college	10	0	Sales	Not-in-family	Asian-Pac-Islander	Female	32	0	United-States	<=50K
3	38	Self-emp-not-inc	44426	HS-grad	9	0	Sales	Unmarried	White	Male	30	0	United-States	>50K
4	62	Without-pay	33230	HS-grad	9	1	Sales	Husband	White	Male	-4	0	United-States	<=50K
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
8114	30	Private	298018	Bachelors	12	0	Adm-clerical	Other-relative	White	Female	20	0	Mexico	<=50K
8115	40	Private	186841	10th	6	0	Handlers-cleaners	Unmarried	White	Male	0	0	United-States	<=50K
8116	44	Private	116414	1st-4th	3	1	Craft-repair	Wife	Other	Female	10	0	Honduras	<=50K
8117	41	Federal-gov	185089	Bachelors	12	0	Machine-op-inspct	Not-in-family	Black	Female	1	0	United-States	<=50K
8118	28	Private	27152	Bachelors	12	0	Exec-managerial	Not-in-family	White	Female	21	0	United-States	<=50K

8119 rows × 14 columns

df_test["marital-status"]=df_test["marital-status"].astype(int)
dataset.dtypes

age               int64
fnlwgt            int64
education-num     int64
marital-status    int32
sex               int32
capital-gain      int64
capital-loss      int64
hours-per-week    int64
income            int32
dtype: object

df_test.drop(labels= ['workclass','education','occupation','relationship','race','native-country'],axis =1, inplace = True)

df_test.income.unique()

array([' <=50K', ' >50K'], dtype=object)

df_test['income']= df_test['income'].map({" <=50K":0, " >50K":1})
display(df_test)

	age	fnlwgt	education-num	marital-status	sex	capital-gain	capital-loss	income
0	40	106899	8	0	Male	0	0	0
1	38	157191	8	1	Male	0	0	0
2	76	102547	10	0	Female	32	0	0
3	38	44426	9	0	Male	30	0	1
4	62	33230	9	1	Male	-4	0	0
...	...	...	...	...	...	...	...	...
8114	30	298018	12	0	Female	20	0	0
8115	40	186841	6	0	Male	0	0	0
8116	44	116414	3	1	Female	10	0	0
8117	41	185089	12	0	Female	1	0	0
8118	28	27152	12	0	Female	21	0	0

8119 rows × 8 columns

df_test['sex']= df_test['sex'].map({" Male":0, " Female":1})
display(df_test)

	age	fnlwgt	education-num	marital-status	sex	capital-gain	capital-loss	income
0	40	106899	8	0	0	0	0	0
1	38	157191	8	1	0	0	0	0
2	76	102547	10	0	1	32	0	0
3	38	44426	9	0	0	30	0	1
4	62	33230	9	1	0	-4	0	0
...	...	...	...	...	...	...	...	...
8114	30	298018	12	0	1	20	0	0
8115	40	186841	6	0	0	0	0	0
8116	44	116414	3	1	1	10	0	0
8117	41	185089	12	0	1	1	0	0
8118	28	27152	12	0	1	21	0	0

8119 rows × 8 columns

df_test.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8119 entries, 0 to 8118
Data columns (total 8 columns):
 #   Column          Non-Null Count  Dtype
---  ------          --------------  -----
 0   age             8119 non-null   int64
 1   fnlwgt          8119 non-null   int64
 2   education-num   8119 non-null   int64
 3   marital-status  8119 non-null   int32
 4   sex             8119 non-null   int64
 5   capital-gain    8119 non-null   int64
 6   capital-loss    8119 non-null   int64
 7   income          8119 non-null   int64
dtypes: int32(1), int64(7)
memory usage: 475.8 KB

from sklearn.preprocessing import MinMaxScaler
min_max_sc = MinMaxScaler()
df_test=min_max_sc.fit_transform(df_test)

df_test_predict=results.predict(df_test)

predictions=df_test_predict

prediction = pd.DataFrame(predictions, columns=['hours-per-week']).to_csv('prediction.csv')